dafny-verification

Stub. Elicit software correctness obligations, maintain a recoverable correctness workpiece, and author or review Dafny specifications with an honest account of what was stated, assumed, discharged, skipped, or trusted. Use for a correctness interview or a Dafny specification or proof review.

Quality

52%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./libs/@hashintel/brunch-agent/packages/plugin-dafny/src/skills/dafny-verification/SKILL.md

Quality

Content

30%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is an explicitly self-declared stub: lean and clearly organized, but with no procedure, no executable guidance, and no workflow. The proposed disclosure shape is sensible, yet every referenced file is absent from the bundle, so even the structural promise is unfulfilled.

Suggestions

Author the actual procedure (elicitation steps, workpiece recording, specification review) or remove the skill until it exists, so the body contains actionable guidance rather than a placeholder.

Add a workflow with explicit validation checkpoints, e.g., run the Dafny verifier on edited specifications and only proceed when verification passes, with a fix-and-revalidate loop.

Create the referenced files (references/dafny-specification.md, references/proof-checks.md, templates/workpiece.md) or drop the paths, so the disclosure tree points at real bundle content.

Dimension	Reasoning	Score
Conciseness	The body is very lean (~20 lines) with no padding and no explanation of concepts Claude already knows. Not 5 because 'the accepted Ampcode pressure test' is unexplained insider context that spends tokens without informing the reader.	4 / 5
Actionability	The body states it 'authors no procedure yet' and only records a proposed file layout; there is no concrete code, command, or instructional guidance. Matches the anchor for entirely vague/abstract content that describes rather than instructs; not 2 because even high-level executable hints are absent.	1 / 5
Workflow Clarity	No sequence of steps exists anywhere; the tree diagram is a proposed structure, not a workflow, and there are no validation checkpoints. Matches the steps-missing anchor; not 2 because there is not even a rough sequence with gaps.	1 / 5
Progressive Disclosure	The body presents a clear one-level-deep tree with a labeled condition ('when recording or revising'), but none of the referenced paths (references/dafny-specification.md, references/proof-checks.md, templates/workpiece.md, the 'elicitation' skill) exist in the bundle, so navigation dead-ends. Not 4 because the pointers are broken; not 2 because the inline structure is short and genuinely well organized.	3 / 5
	Total	9 / 20 Passed

Description

75%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, has explicit and reasonably natural triggers, and answers both what and when in third person. It is held back from the top level by abstract jargon ('recoverable correctness workpiece'), a when-clause narrower than its what-clause, and missing common synonyms like 'formal verification' or '.dfy'.

Suggestions

Add natural trigger synonyms such as 'formal verification', 'proof assistant', and '.dfy files' to the Use-for clause so it fires on how users actually phrase Dafny requests.

Replace or gloss the abstract phrase 'recoverable correctness workpiece' with a concrete capability statement (e.g., 'record and revise correctness obligations and their discharge status in a workpiece file').

Extend the when-clause to cover the authoring use case (e.g., 'Use when writing or reviewing Dafny specifications...') so the triggers match the full set of stated capabilities.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('Elicit software correctness obligations', 'maintain a recoverable correctness workpiece', 'author or review Dafny specifications') with an honest-accounting clause, matching the several-specific-actions anchor. Not 5 because 'recoverable correctness workpiece' is abstract jargon rather than a fully concrete capability, and not 3 because more than 1-2 specific actions are named.	4 / 5
Completeness	Both what (elicit, maintain, author or review) and when ('Use for a correctness interview or a Dafny specification or proof review') are explicitly present. Not 5 because the when-clause covers only interview/review triggers while the what-clause also claims authoring, so the when could be more specific.	4 / 5
Trigger Term Quality	'correctness interview', 'Dafny specification', and 'proof review' are natural phrases a user would say, giving good keyword coverage. Not 5 because common synonyms like 'formal verification', 'proof assistant', or '.dfy' are missing; not 3 because the key natural trigger phrases are present rather than only domain labels.	4 / 5
Distinctiveness Conflict Risk	'Dafny' and 'correctness interview' give a clear niche marker with mostly distinct triggers. Not 5 because the non-Dafny clauses ('elicit software correctness obligations', 'honest account of what was stated, assumed...') are broad and could overlap with general correctness-review or elicitation skills.	4 / 5
	Total	16 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 3 missing	Warning

	Total	15 / 16 Passed

Repository: hashintel/hash
Path: libs/@hashintel/brunch-agent/packages/plugin-dafny/src/skills/dafny-verification/SKILL.md
Commit: f9f024c

Reviewed: 10 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.